There is a need of providing sensors in mobile terminals in order to make the mobile terminal capable of sensing its ambient conditions. The information can be used for context awareness applications where the ambient information is used for controlling the user interface profile and different settings of the mobile terminal user interface. Fingerprint sensors are also needed for authenticating the user of the terminal.
There exist several kinds of fingerprint sensors: skin impedance based sensor, thermal sensors, and optical sensors. The most practical solution for authentication of a user of small appliances, such as mobile terminals, is based on capacitive impedance measurement. The basic idea of the capacitive fingerprint sensor to measure the change of skin impedance is described in FIGS. 1A and 1B. An array of sensors 120 measure the skin impedance values when a finger 101 is gradually pulled over the array of sensors. The capacitance between the electrode surface and the conductive saline layer inside the skin surface varies as a function of distance to the conductive layer. The varying air gap and the dead horny cells in the surface of the skin form the capacitance 125 to the conductive layers 121, 122 forming the electrodes of the capacitive sensor.
FIG. 2 shows a rough equivalent circuit of the skin impedance and the impedance measurement principle. Skin has a fixed resistive tissue component 202, and a fixed resistive surface component 203. The measurement capacitance also has a fixed component 272 and a component 271 that varies according to the surface form of the finger. The capacitive fingerprint sensor measures the varying capacitive component by applying an alternating voltage 281 to a drive electrode 222 and measuring a signal from a sensor electrode 221. The signal is amplified with a low noise amplifier 282, and the phase difference between driver and sensing electrodes is measured, 283. Interference can be suppressed with a guard electrode, which is kept in the same potential as the signal input using a buffer 285.
A fingerprint sensor also requires a signal processing circuit, which is preferably a silicon-based integrated circuit. One solution for providing a fingerprint sensor would be to use an integrated circuit, which would serve both as capacitive measurement electrodes and as signal processing electronics. This integrated circuit would then be mounted on the surface of the appliance. However, the area needed for the capturing the capacitive image of the fingerprint is roughly in the scale of one square centimeter. This is a very large area for using a silicon integrated circuit as measurement electrodes. Furthermore, the measurement consists of hundreds of capacitive pixels that are arranged in a row or in a matrix depending on the measurement principle. A lot of wiring is needed and the measurement electrodes need to be isolated from the integrated circuits. Therefore a cost efficient method for connecting the capacitive electrodes to the signal processing silicon integrated circuit is needed.
One typical prior art solution is described in patent documents U.S. Pat. No. 5,887,343 and U.S. Pat. No. 6,067,368. The problem is solved by using a separate insulating planar substrate to form the measurement electrode. This substrate contains the interconnecting wiring and the vias through the substrate. The substrate is connected to the silicon integrated circuit containing the signal and data processing capabilities. However, this solution is complicated to manufacture because a large number of interconnecting wiring must be connected within a small space. Such wiring also is not very robust, which tends to make the structure easy to break in mobile use.
Another prior art solution is to create the measurement electrodes directly on top of the silicon wafer. This leads to a simple configuration of interconnecting wiring but the solution requires a large silicon surface due to the large area needed for the electrodes.
The prior art solutions also have a disadvantage that relates to security. It is possible to make external connections to the wiring between the capacitive measurement electrodes and the integrated circuit, and by using such a connection it is possible to “record” signals that relate to a certain finger when the finger is measured. It is then later possible to input these recorded signals to the integrated circuit and thus a positive authentication result can be achieved electrically without any finger.
A further disadvantage with the prior art solutions relates to the ergonomics of the sensor. A finger must be pressed rather heavily against the flat sensor in order to achieve sufficient contact area between the sensor and the finger. Therefore the measurement may often fail when the finger is not pressed and slid properly along the sensor surface.